0

For a project, I have a csv file with 60 coordinates and the radius (for each centroid of a city district) I want to get my Google Maps results of. Aim is to loop through the coordinates and the radius and receive all existing places of one certain type (e.g. cafes) for each of the listed coordinates. So far, I was able to retrieve the results with adding the coordinates manually and I'm struggling with applying the pandas 'apply' to my 'givePlaces'-function.

I used pandas to import the csv file and wrote a function that returns me the max. possible results (<60, Google doesn't allow more). I defined the variables "location", "radius" and "type". The code looks the following and almost works. Maybe one of you can help:

import googlemaps
import pprint
import time
import csv
import pandas as pd
from GoogleMapsAPIKey import get_my_key 

#import the coordinates and radius csv-file
df = pd.read_csv('filepath',sep=';',header=0)
df

Simplified, the csv-table looks the like the following:

Latitude Longitude Radius
48.179037 11.522759 500
48.156327 11.414132 1200 
#define my API Key
API_KEY = get_my_key()
gmaps = googlemaps.Client(key = API_KEY)

#Define a function that returns the Places API results:
def givePlaces(location, radius, type):
    # define list of places of interest
    places = []
    # request first page
    places_result = gmaps.places_nearby(location=location, radius=radius, type=type)
    # add results to places
    places = places + places_result['results']
    # while 'next_page_token' exists, request the next page and repeat
    while('next_page_token' in places_result):
        # wait 3 seconds before next request
        time.sleep(3)
        # overwrite places_result with results of second and/or third page
        places_result = gmaps.places_nearby(page_token = places_result['next_page_token'])
        # add these results to places
        places = places + places_result['results']
    print("found " + str(len(places)) + " places")
    return places

#Previously I used this manually:
location = 48.179037,11.522759
radius = 500
type = 'cafe'
my_places = givePlaces(location, radius, type)

#the apply function would look like this, but so far does not work:
df['places'] = df.apply(lambda row: givePlaces((row['Länge'], row['Breite'], row['Radius'], 'cafe'), axis=1)

pprint.pprint(my_places)

Can anyone kindly assist me with the pandas apply function? I'm still quite a beginner in coding and Google did not help much in my specific issue. Thank you so much in advance!

marinade
  • 5
  • 2
  • You can find some comprehensive description on iterating pandas data frames https://stackoverflow.com/questions/16476924/how-to-iterate-over-rows-in-a-dataframe-in-pandas however I would advise to re-think (and maybe redefine) asking the question after reading above mentioned as it can be matched as a duplicate. – Kube Kubow Nov 14 '21 at 20:16
  • 1
    Don't loop, use [apply](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.apply.html): `df['places'] = df.apply(lambda row: givePlaces((row['Latitude'], row['Longitude'), row['Radius'], 'cafe'), axis=1)` – RJ Adriaansen Nov 14 '21 at 20:20
  • @RJAdriaansen thank you, I tried it. Unfortunately it returns "Unexpected EOF while parsing". – marinade Nov 14 '21 at 20:36
  • @KubeKubow thank you, good idea! Your comment and the one from RJ Adriaansen helped me to redefine the question. I think using 'apply' could definitely solve the problem :). Still struggling with implementation though... – marinade Nov 14 '21 at 20:54

1 Answers1

0

3 things:

  • don't name a variable type, as it is also a built-in python function
  • df.apply(lambda row: givePlaces((row['Länge'], row['Breite'], row['Radius'], 'cafe'), axis=1) is incorrectly formatted, it's missing a parenthesis and you need to pass location, as a string, dict, list or tuple.
  • The location needs to be specified as latitude (Breite),longitude (Länge)

Using a tuple that would make:

df.apply(lambda row: givePlaces((row['Breite'], row['Länge']), row['Radius'], 'cafe'), axis=1)
RJ Adriaansen
  • 9,131
  • 2
  • 12
  • 26